Browser MCP is an MCP server plus browser integration that lets AI apps automate the user’s real browser instead of launching a separate automation session. By adapting a Playwright-style approach to control the running browser profile, it reuses logged-in sessions and cookies, which reduces re-authentication friction and helps avoid some bot-detection heuristics. The server exposes structured tools for navigation, element interaction, and artifact capture (DOM, screenshots, logs), all discoverable via MCP schemas. Because it runs against the user’s primary browser, it’s well-suited to repetitive web tasks, authenticated dashboards, and debugging workflows inside MCP-capable IDEs. A public website and extension streamline installation and connect the local server to clients like Claude, Cursor, VS Code, and Windsurf. The repository shows active development and a growing star count, reflecting rapid adoption across agent tooling.
Features
- Control of the user’s existing browser profile
- Capture of DOM, console/network logs, and screenshots
- Navigation, form fill, click, and element interaction tools
- Works with Claude, Cursor, VS Code, and Windsurf
- MCP schemas for typed, discoverable automation
- Session reuse to reduce logins and mitigate detection